Tottenham and Chelsea are in a slightly false position in the Premier League according to the underlying numbers.
If we use Opta’s xPts model, Tottenham would be way down in 13th place, while Chelsea and Bournemouth would be in mid-table.
For comparison, Arsenal are exactly where they should be in first place – they clearly are the Premier League’s best side – and Liverpool are in line with their underlying numbers in fourth.
The only other clubs in the ‘correct’ position are Everton and West Ham.
What are XPts and how are they calculated?
According to football analytics expert McKay Johns:
In a basic approach, xPts is calculated by simulating each shot in a match.
For each shot, the model will generate a random number between 0 and 1. If the number is less than the value of the xG of that shot, it says it is a goal and adds to that game’s simulation tally, and at the end compares the actual goal tallies.
If a team wins that simulated match they get 3 points, draw they get 1 point, and a loss they get zero.
Running that same simulation over many iterations (say 1000) gives the output of how often a team would have won, drawn, or lost, and then we can take the average of all the points they have accumulated in the simulations to get their expected points which will fall between 0 and 3.
So basically, it’s running the game through algorithms and working out the average of how many points should have been garnered from each match according to the xG created and the xG conceded.
As Johns says: ‘The stat does well at giving us a decent idea of how many points a team should get on average, especially over a season.’
The Premier League’s biggest overperformers:
5) Chelsea – overperforming by five places
It’s tricky to know what to make of Chelsea this season; they do not feel like a functional football team and yet they sit just two points behind second-placed Manchester City in fifth and have won their last two Premier League games. But an xPts calculation puts them in tenth, largely because they are conceding better quality chances than bottom club Wolves and indeed every other team barring Manchester United, West Ham, Nottingham Forest and Burnley.
And at the other end, Moises Caicedo is the the Premier League player who is most exceeding his individual xG with actual goals.
Is Enzo Maresca any good? We really do not have an answer.
4) Bournemouth – overperforming by six places
Andoni Iraola came out top in our manager rankings for the season as he somehow has Bournemouth in third place. But the underlying numbers would have them down in ninth as they should have conceded marginally more goals than they have scored. It seems like their forced defensive overhaul of the summer could have come at a hefty price, were it not for a decent goalkeeper and the astonishing form of Antoine Semenyo.
3) Aston Villa – overperforming by seven places
The underlying numbers would put Villa bang in a relegation battle; astonishingly, only Burnley are struggling more to create good chances. Indeed, in terms of creating actual Big Chances, only Burnley and Wolves are worse. They were lucky to claim draws from Newcastle and Everton when barn doors were being missed by the opposition, while they were gifted a win by Tottenham last time out.
And somehow they have clawed their way into mid-table. Could this be the season they end their ridiculous trophy drought?
2) Sunderland – overperforming by seven places
A phenomenal job has been done by Sunderland this season, but having a brilliant goalkeeper has certainly helped them massively under-concede, leaving them far advanced on the 14th place their xPts for the season would determine. They are unbeaten at home without ever really dominating any game, while they were one of several teams to benefit from the profligacy of Crystal Palace.
But nobody would surely begrudge them their place in the higher echelons of the Premier League; their recruitment has been exemplary.
1) Tottenham – overperforming by seven places
We asked last week about the actual point of Spurs because they do appear to be something of a muddled mess. They should seemingly be 13th rather than sixth, with even the mess of Nottingham Forest creating better chances across the season. But they are the team most over-performing on their xG, with Joao Palhinha, Richarlison and Brennan Johnson all far exceeding their individual expected numbers this season.
It’s telling that we would absolutely have guessed that Spurs would be top of this list…they just don’t feel like a team who should be just a point behind Liverpool.
The full Premier League xPts table:
1) Arsenal
2) Crystal Palace
3) Manchester City
4) Liverpool
5) Newcastle United
6) Manchester United
7) Leeds United
8) Brighton
9) Bournemouth
10) Chelsea
11) Brentford
12) Everton
13) Tottenham
14) Sunderland
15) Wolves
16) Fulham
17) Nottingham Forest
18) Aston Villa
19) West Ham
20) Burnley
